DESCRIPTION (Applicant's Description): Evidence from ecological, epidemiological and laboratory studies consistently supports the contention that an overall pattern of dietary consumption can provide a protective effect against breast cancer and its recurrence. This protective pattern is characterized by a high daily consumption of vegetables, fruits and fiber with low levels of energy consumed from fat. The introduction of such a diet after breast cancer diagnosis may reduce the recurrence rate by 30-40 percent. In a preliminary randomized feasibility study we have demonstrated that breast cancer survivors can be motivated through a telephone counseling program to habituate to such daily dietary pattern that includes 5 Vegetable servings, 16 ounces of vegetable juice, 3 fruits, 30 grams of fiber with 15-20 percent of energy consumed from fat. With University gift monies, we have started a multicenter randomized controlled trial to test whether the adoption and maintenance of this diet reduces recurrence in breast cancer survivors from selected regions of California, Arizona, and Texas. More than 600 women have been randomized out of a target of 3,000 who are within 4 years of diagnosis of stage I (greater than or equal to 1 cm), stage II and stage IIIA disease and who have completed conventional therapy. To date we have demonstrated that we are able to maintain adequate adherence to the dietary pattern in 89 percent of women who have completed 12 months on this study while only 8 percent of the control group were adherent. WE validate adherence with circulating carotenoid and estrogen biomarkers. Participants are contacted every 6 months for a health status assessment and medical records are reviewed for all reported cancers or deaths. To gain insight into the mechanism of how diet might influence the disease, we will compare circulating carotenoid and estrogen levels in women who have a secondary cancer event with levels in matched set of women who have not recurred. In addition, we are collecting and storing buffy coat as well as tumor specimens for future correlative studies. If our study groups have prognoses similar to women in recently published studies, we will have greater than 0.90 power to detect a 16 percent improvement in the breast cancer events in the study period. Regardless of the outcome of the dietary intervention, the results of this study will have major implications for future directions of research in breast cancer control.